Is a locked tag really safe from being overwritten?
The bytes are genuinely safe: a locked tag cannot be rewritten by anyone. The object is not. It can be covered with another tag, prised off and swapped, or simply removed, and none of those attacks touch the memory locking protects.
What locking genuinely protects
The content, completely. Nobody can rewrite a locked tag, and that is not a probabilistic statement: the chip will not accept the write.
For the specific threat of a passer by pointing your poster somewhere else with a phone, locking is a complete answer. That threat is real and locking removes it.
What it leaves open
Everything physical, and physical is where the easier attacks live.
A second tag placed over the first, which the phone will read instead. Removal and replacement with a tag holding different content. Simple destruction, which does not redirect anyone and does defeat the deployment.
None of these require any skill with NFC, and all of them are cheaper for an attacker than rewriting a tag would have been.
The cover attack
Worth examining because it is the one people do not anticipate and the one that actually happens.
Sticking a new tag over a locked one gives a phone two tags in the field, and the outcome depends on geometry. Often the outer tag wins because it is closer. Sometimes both fail because they detune each other, which is at least a visible failure rather than a silent redirection.
The same pattern is well established with printed codes, where covering one with another is a common fraud. NFC is not immune to it, and locking does nothing about it.
What real protection looks like
Physical, first. A tag under a laminate, behind a fixed panel, embedded in an object or built into a card is far harder to cover or swap than a sticker on a flat surface.
Visible, second. A design where the tag's location is obvious and a foreign sticker looks wrong is worth more than any chip feature, because it recruits every passer by as an observer.
Cryptographic, where the stakes justify it. A chip that authenticates rather than merely holds content lets your server verify that the tag is the genuine one, which is the only mechanism that survives a substitution.
Locking belongs alongside these rather than instead of them. It is one layer, it costs nothing, and treating it as sufficient is where deployments get caught.
Frequently asked questions
Can someone stick another tag over mine?
Yes, and it is the most likely attack. Whether the outer tag wins depends on geometry, and both failing is also a common outcome since two coils in contact interfere.
Does locking prevent cloning?
No. Anyone can read a locked tag and write the same content to a blank one. Locking prevents modification of that tag rather than duplication of what it says.
How do I detect that a tag was replaced?
By checking the identifier against a record of what you deployed, or by using a chip that authenticates. Content alone cannot tell you, since a copy holds identical content.
Is a tamper evident tag worth it?
Where removal is the threat, yes. Those tags break their antenna when peeled, so a removed tag stops working rather than being reusable elsewhere.